Tailscale fixed flaws in its built-in SSH feature that could let one hacked device take root control of other servers. Self-hosted headscale users are still exposed.
Tailscale has released version 1.104.4 to fix flaws in its built-in SSH server. The flaws could let an attacker who already controls one trusted device reach files and services on another server that only the all-powerful root user should be able to touch. Anyone running Tailscale SSH on a server should update.
Tailscale is a tool that links computers and servers into a private network. Its SSH feature lets people open remote sessions without passwords or keys. Access depends on the device's identity and the network's rules. The flaws are in how that feature forwards connections. The Tailscale background process does this job with root privileges, and a timing trick let an attacker slip past the safety check.
Security researcher Erica Windisch of Cyberdione Labs published the findings in a report called Tailworm. In her lab tests, forwarding to the Docker control socket gave a root shell and let her read the system password file.
The server side of Tailscale SSH runs only on Linux and the open-source macOS build. The attacker's device can be any platform. An attacker first needs control of a device the network rules already trust to open SSH sessions, with forwarding allowed. Tailscale says this does not let an unknown outsider connect to random servers.
People who run headscale, the self-hosted control server, are in a worse spot. Version 0.29.5 does not fix the reported issues. Windisch says every SSH grant there should be treated as giving root.
The alert box above lists the steps, including a workaround if you cannot update yet. One point deserves attention: if a server may already be compromised, updating is not enough. Windisch says a stolen device key stays valid until the device is deleted from the control console. Tailscale's notices do not cover this case.
Windisch called the chain wormable, saying one device could spread to every device the rules allowed. Tailscale rejects that label and describes movement limited to connections the rules already permit. It calls the problems serious. The four bulletins have no CVE numbers, no severity scores and no oldest affected version. Neither side says the flaws have been used in real attacks, and the proof-of-concept code is not public.
Update every server that uses Tailscale SSH to version 1.104.4 or later, and delete any device you think was compromised.
A user says Anthropic's Claude Code put his email in web requests, then scam messages arrived. The link is unproven, but it shows what AI tools can share.
Read more: User says Claude Code sent his email address to websitesA viral ChatGPT prompt turns what the chatbot remembers about you into a room image. Posting it publicly could give scammers clues to target you.
Read more: Viral ChatGPT room image could hand scammers clues about youA message saying your Amazon account was hacked is often the scam itself. Checking in the Amazon app first can protect your card, gift balance and phone.
Read more: If Amazon says your account was hacked, check before you call
